that if we had been accompanied by a puppy dog, in addition to the artist, the dog would have acted immediately on the hypothesis of a chair and would have jumped onto it by way of using it as such. Again, if the dog had refrained from such action, it would have been because it was a well-trained dog. Therefore the transition from a coloured shape to the notion of an object which can be used for all sorts of purposes which have nothing to do with colour, seems to be a very natural one; and we—men and puppy dogs—require careful training if we are to refrain from acting upon it.
Thus coloured shapes seem to be symbols for some other elements in our experience, and when we see the coloured shapes we adjust our actions towards those other elements. This symbolism from our senses to the bodies symbolized is often mistaken. A cunning adjustment of lights and mirrors may completely deceive us; and even when we are not deceived, we only save ourselves by an effort. Symbolism from sense-presentation to physical bodies is the most natural and widespread of all symbolic modes. It is not a mere tropism, or automatic turning towards, because both men and puppies often disregard chairs when they see
